A data scientist has ground truth labels for a dataset and wants to evaluate the clustering results from K-Means. They compute the adjusted Rand index (ARI) and obtain a score of 0.8. What does this score indicate?

The clustering is highly similar to the ground truth, with 0.8 indicating strong agreement beyond chance.

The clustering is worse than random, because ARI should be above 1.

The clustering is only 80% accurate, meaning 20% of points are misclustered.

The clustering is not reliable because ARI is sensitive to the number of clusters.

A team is clustering sensor readings from a manufacturing process. The data contains normal operating states and rare anomalous events. They want to identify the normal states as clusters and treat the anomalies as noise. Which algorithm is most suitable?

Gaussian Mixture Model, because it provides probabilities.

Agglomerative hierarchical clustering, because it produces a dendrogram.

K-Means, because it is simple and fast.

DBSCAN, because it can label low-density regions as noise.

A data scientist is clustering text documents based on word frequencies. They have a high-dimensional sparse matrix. They want to use K-Means but are concerned about the curse of dimensionality. Which distance metric is most appropriate for high-dimensional sparse text data?

Cosine similarity, because it measures the angle between vectors and is robust to the magnitude of word counts.

Euclidean distance, because it is the default for K-Means.

Manhattan distance, because it is less sensitive to outliers.

Jaccard distance, because it is designed for binary data.

A data engineer is clustering a dataset with 1000 features and 10,000 samples. They plan to use K-Means. They are concerned about the curse of dimensionality. Which approach is most effective?

Apply PCA to reduce the dimensionality to 50 components before clustering.

Increase the number of clusters to account for the high dimensionality.

Use a distance metric that is immune to the curse of dimensionality, such as Manhattan distance.

Use K-Means directly on the original features because it is robust to high dimensionality.

GIAC Machine Learning Engineer

Domain 2Objective 1

Clustering GMLE Practice Questions (Page 9)

Part of the Machine Learning Algorithms domain, which makes up ~29% of our current practice bank. GIAC (SANS) does not publish an official question count, but from its 180-minute exam (~70–120 total, ~20–35 in this domain), expect 7–12 from this objective — we provide 45 practice questions to prepare you well beyond it. (estimate)

45questions here

9free pages

8concepts

Questions 41–45

41application · medium
A research team is clustering documents by topic. They have a term-document matrix with TF-IDF weights. They want to use hierarchical clustering and need a distance metric that is appropriate for high-dimensional sparse text data. Which distance metric should they use?
Select an answer first
42expert · hard
A data scientist has ground truth labels for a dataset and wants to evaluate the clustering results from K-Means. They compute the adjusted Rand index (ARI) and obtain a score of 0.8. What does this score indicate?
Select an answer first
43application · medium
A team is clustering sensor readings from a manufacturing process. The data contains normal operating states and rare anomalous events. They want to identify the normal states as clusters and treat the anomalies as noise. Which algorithm is most suitable?
Select an answer first
44application · medium
A data scientist is clustering text documents based on word frequencies. They have a high-dimensional sparse matrix. They want to use K-Means but are concerned about the curse of dimensionality. Which distance metric is most appropriate for high-dimensional sparse text data?
Select an answer first
45expert · hard
A data engineer is clustering a dataset with 1000 features and 10,000 samples. They plan to use K-Means. They are concerned about the curse of dimensionality. Which approach is most effective?
Select an answer first

Finished these 5 questions?

Review the revealed explanations, or continue through the curriculum.

No more pagesBack to GMLE

Free Basic Practice is a study aid with revealable answers — not a scored exam. Examers.io is independent and not affiliated with or endorsed by GIAC (SANS). “GMLE” is a trademark of its owner, used for identification only.